Replication Protein A as a “Fidelity Clamp” for DNA Polymerase α*

The current view of DNA replication in eukaryotes predicts that DNA polymerase α (pol α)-primase synthesizes the first 10-ribonucleotide-long RNA primer on the leading strand and at the beginning of each Okazaki fragment on the lagging strand. Subsequently, pol α elongates such an RNA primer by incorporating about 20 deoxynucleotides. pol α displays a low processivity and, because of the lack of an intrinsic or associated 3′→ 5′ exonuclease activity, it is more error-prone than other replicative pols. Synthesis of the RNA/DNA primer catalyzed by pol α-primase is a critical step in the initiation of DNA synthesis, but little is known about the role of the DNA replication accessory proteins in its regulation. In this paper we provide evidences that the single-stranded DNA-binding protein, replication protein A (RP-A), acts as an auxiliary factor for pol α playing a dual role: (i) it stabilizes the pol α/primer complex, thus acting as a pol clamp; and (ii) it significantly reduces the misincorporation efficiency by pol α. Based on these results, we propose a hypothetical model in which RP-A is involved in the regulation of the early events of DNA synthesis by acting as a “fidelity clamp” for pol α.

The current view of DNA replication in eukaryotes predicts that DNA polymerase ␣ (pol ␣)-primase synthesizes the first 10-ribonucleotide-long RNA primer on the leading strand and at the beginning of each Okazaki fragment on the lagging strand. Subsequently, pol ␣ elongates such an RNA primer by incorporating about 20 deoxynucleotides. pol ␣ displays a low processivity and, because of the lack of an intrinsic or associated 33 5 exonuclease activity, it is more error-prone than other replicative pols. Synthesis of the RNA/DNA primer catalyzed by pol ␣-primase is a critical step in the initiation of DNA synthesis, but little is known about the role of the DNA replication accessory proteins in its regulation. In this paper we provide evidences that the singlestranded DNA-binding protein, replication protein A (RP-A), acts as an auxiliary factor for pol ␣ playing a dual role: (i) it stabilizes the pol ␣/primer complex, thus acting as a pol clamp; and (ii) it significantly reduces the misincorporation efficiency by pol ␣. Based on these results, we propose a hypothetical model in which RP-A is involved in the regulation of the early events of DNA synthesis by acting as a "fidelity clamp" for pol ␣.
The highly conserved DNA polymerase (pol) 1 ␣-primase complex is the only eukaryotic polymerase able to initiate DNA synthesis de novo, making it of central importance in DNA replication. In fact, it is required both for the initiation of DNA replication at chromosomal origins and for the discontinuous synthesis of Okazaki fragments on the lagging strand of the replication fork (1)(2)(3). The current view of DNA replication in eukaryotes predicts that pol ␣-primase synthesizes the first RNA/DNA primer on the leading strand. Then, at a critical length of 30 nucleotides, replication factor C binds to the 3Ј-OH end of the nascent DNA strand and displaces pol ␣, thereby loading PCNA and pol ␦. pol ␣-primase switches activity to initiate the synthesis of Okazaki fragments on the lagging strand (4 -6). pol ␣-primase has to synthesize short RNA/DNA primers. Thus, its intrinsic low processivity is compatible with its function. However, the switch between primase and polym-erase activity, leading to the RNA-to-DNA synthesis transition, occurs through an intramolecular mechanism that does not require dissociation of pol ␣ from the primer (7). Thus, it would be important to ensure stable binding of pol ␣ to the template until completion of the RNA/DNA primer synthesis. In addition, the lack of any intrinsic proofreading function for pol ␣ could lead to misincorporation events during RNA/DNA primer synthesis. The critical roles of the pol ␣-primase make it a likely target for mechanisms that control DNA synthesis initiation and progression (8). In particular, a mechanism ensuring a transient but stable binding of pol ␣ to the primer and increasing its fidelity could represent an advantage for the cell.
The eukaryotic ssDNA-binding protein RP-A is a heterotrimer consisting of three subunits of 70, 32, and 14 kDa (p70, p32, and p14, respectively) (9 -11). RP-A has multiple roles in the cell, being essential for DNA replication initiation and elongation (12,13), DNA repair (14), and DNA recombination (15). These different roles are mediated by direct protein/protein interactions. RP-A has been found to interact physically with the replicative OBP (origin-binding protein)/helicase large T-antigen of SV40 (16 -18), with pol ␣-primase (19), with proteins of the nucleotide excision repair (NER) pathway (XPA, XPG, XPF) (20 -24), with recombination-specific proteins (15,25) and with transcriptional activators (GAL4, VP16, p53, RBT1) (26 -30). The three subunits of RP-A are likely to play different roles in all these DNA transactions. In particular, mutagenesis studies have begun to reveal different functions of p70, p32, and p14 in DNA replication. All three subunits are required to support DNA replication in vitro. The p70 subunit contains three functionally distinct domains: an N-terminal domain, a central ssDNA-binding domain, and a C-terminal subunit interaction domain (31). The N-terminal domain contains the interaction domain for pol ␣-primase, which consists of two distinct regions: one (amino acids 1-170) that stimulates pol ␣ synthetic activity and another (amino acids 170 -327) that increases pol ␣ processivity. This latter region overlaps with the ssDNA-binding domain (amino acids 168 -450); such an activity was shown to be required for pol ␣ processivity stimulation by RP-A (19). The p32 subunit interacts with XPA and large T-antigen and is phosphorylated in a cell cycle-dependent manner (32). UV-cross-linking studies with photoreactive primer-templates mapped both p70 and p32 close to the 3Ј-OH primer end (33)(34)(35)(36). The pattern of p70 and p32 labeling was also found to be dependent on the template length and the ratio of RP-A to template concentration. These studies suggested that a series of conformational changes occurs after RP-A binding to ssDNA. Indeed, at least two different modes of RP-A binding to DNA have been detected: one covering 8 -10 nucleotides and another with a more extended configuration of 30 nucleotides (37)(38)(39)(40). These structurally distinct complexes have been proposed to have different subunits rearrangements.
Very recently, it has further been demonstrated that RP-A, along with the DNA replication protein Cdc45p, is involved in the recruitment of pol ␣-primase at the chromosomal DNA replication origins (41,42). Thus, both pol ␣-primase and RP-A are simultaneously present and likely to interact physically during initiation of DNA replication. This makes RP-A a likely candidate for the regulation of the catalytic activity of pol ␣-primase.
In the present work, we have investigated the in vitro role of RP-A on the DNA synthetic activity of pol ␣ using different DNA templates. The results obtained indicated that RP-A could assist pol ␣ in two ways: (i) by increasing the stability of the pol/primer complex and (ii) by reducing the overall misincorporation rate of pol ␣.
Nucleic Acid Substrates-The singly primed d24:d66-mer was prepared by labeling the 5Ј-end of the d24-mer primer with [␥-32 P]ATP and T4 polynucleotide kinase (Ambion) according to the manufacturer's protocol. The d66-mer template oligonucleotide was then mixed with the complementary labeled d24-mer primer oligonucleotide in a 1:1 molar ratio in 20 mM Tris-HCl (pH 8.0) containing 20 mM KCl and 1 mM EDTA, heated at 90°C for 5 min, and then incubated at 65°C for 2 h and slowly cooled at room temperature.
Enzymes and Proteins-Calf thymus pol ␣ was purified as described (54). The pol ␣ used in this study was 250 units/ml (0.2 mg/ml). 1 unit of pol activity corresponds to the incorporation of 1 nmol of total dTMP into acid-precipitable material for 60 min at 37°C in a standard assay containing 0.5 g (nucleotides) of poly(dA)/oligo(dT) 10:1 and 20 M dTTP. Recombinant human RP-A was isolated as described (55).
Enzymatic Assays-pol ␣ activity on poly(dA)/oligo(dT) 10:1 was assayed in a final volume of 25 l containing 50 mM Tris-HCl (pH 7.6), 0.25 mg/ml bovine serum albumin, 1 mM dithiothreitol, 6 mM MgCl 2, and 5 M [ 3 H]dTTP (5 Ci/mmol), unless otherwise indicated in the figure legends. All reactions were incubated for 15 min at 37°C unless otherwise stated, and the DNA was precipitated with 10% trichloroacetic acid. Insoluble radioactive material was determined as described (56). When the singly primed d24:d66-mer oligodeoxynucleotide was used as template, a final volume of 25 l contained 50 mM Tris-HCl (pH 7.6), 0.25 mg/ml bovine serum albumin, 1 mM dithiothreitol, 6 mM MgCl 2, and 10 M each [ 3 H]dATP (5 Ci/mmol), dGTP, dCTP, and [ 3 H]dTTP (5 Ci/mmol). Enzymes and proteins were added as indicated in the figure legends. All reactions were incubated for 15 min at 37°C unless otherwise stated and stopped by the addition of 0.1 M EDTA and 1 g of calf thymus DNA as carrier. 20 l of the reaction mixture were then spotted onto DE-81 cellulose filters. Filters were washed to remove unincorporated dNTPs as described (57), and incorporated radioactivity was monitored by scintillation counting.
For incorporation studies with the singly primed d24:d66-mer oligodeoxynucleotide as template, a final volume of 10 l contained 50 mM Tris-HCl (pH 7.6), 0.25 mg/ml bovine serum albumin, 1 mM dithiothreitol, 6 mM MgCl 2 , and 20 nM (3Ј-OH ends) of the 5Ј 32 P-labeled d24:d66mer DNA template. Enzymes, proteins, and unlabeled dNTPs were added as indicated in the figure legends. All reactions were incubated for 15 min at 37°C, samples were mixed with denaturing gel loading buffer (95% v/v formamide, 10 mM EDTA, 0.25 mg/ml bromphenol blue, 0.25 mg/ml xylene cyanol), heated at 95°C for 5 min, and then subjected to electrophoresis on a 7 M urea, 20% polyacrylamide gel. Quantification of the reaction products on the gel was performed using a Molecular Dynamics PhosphorImager and ImageQuant software.
Steady-state Kinetic Data Analysis-K m , V max , and [RP-A] 50 values were calculated according to the Michaelis-Menten equation in the form, where k cat E 0 ϭ V max . The K i values for incorrect dNTPs were determined from inhibition assays with increasing concentrations of the selected dNTP in the presence of different fixed amounts of RP-A and were calculated according to a simple competitive mechanism of inhibition as described by the equation, Computer fitting of the experimental data to the equations was performed with the program MacCurveFit TM 1.5 using the least squares curve-fitting quasi-Newton method, based on the Davidon-Fletcher-Powell algorithm (59). When data points were derived from densitometric analysis of the intensities of the products bands, the values of integrated gel band intensities in dependence of the nucleotide substrate concentrations were fitted to the equation (58), where, T ϭ target site, the template position of interest; and I*T ϭ the sum of the integrated intensities at positions T, Tϩ1 . . . Tϩn. Before being inserted in the above equation, the intensities of the single bands of interest were first normalized by dividing for the total intensity of the lane. This was done to reduce the variability because of manual gel loading. An empty portion of the gel was scanned, and the resulting value was subtracted as background. The goodness of the  [12][13][14][15][16][17][18] (circles) or the heteropolymeric oligodeoxynucleotide d24:d66-mer (squares). B, DNA synthesis by pol ␣ was measured in the presence of increasing amounts of RP-A (as trimer) in the presence of 25 nM (3Ј-OH ends) (squares), 75 nM (3Ј-OH ends) (triangles), or 125 nM (3Ј-OH ends) (circles) of poly(dA)/oligo(dT) [12][13][14][15][16][17][18] . The arrows indicate maximal stimulation at equimolar RP-A to 3Ј-OH primer concentrations. interpolated curve was assessed by computer-aided calculation of the sum of squares of errors and the correlation coefficient R 2 . Standard errors were provided by the computer program MacCurveFit TM 1.5. The standard errors are calculated from the variance-covariance matrix, and the values displayed are the square roots of the diagonal elements. The variance-covariance matrix is calculated from the Jacobian matrix (59).

RP-A Stimulates the Synthetic Activity of pol ␣ in Dependence of the 3Ј-OH Primer
Concentration-Different concentrations of RP-A were tested for their effect on nucleotide incorporation catalyzed by pol ␣ on different DNA templates (Fig. 1). RP-A was able to stimulate pol ␣ activity on either homo-or heteropolymeric deoxyoligonucleotides, within a range of concentrations close to the 3Ј-OH primer concentration used in the assay (Fig. 1A). Variation of the 3Ј-OH primer concentration resulted in a shift of the RP-A concentration giving the maximal stimulation, which was observed at equimolar amounts of RP-A to 3Ј-OH primer (Fig. 1B, arrows).
RP-A Increases the Affinity of pol ␣ for 3Ј-OH Primers-To investigate the effects of RP-A on the nucleotide incorporation reaction catalyzed by pol ␣, increasing concentrations of 3Ј-OH primer were tested in the presence of different fixed amounts of RP-A. As shown in Fig. 2A, the K m value of pol ␣ for the 3Ј-OH primer was decreased by increasing concentrations of RP-A. The V max /K m ratio, which is an estimate of the association rate for the pol ␣⅐3Ј-OH primer complex formation, was also increased by RP-A (Fig. 2B). A comparison of the variation of the reaction velocity in dependence of the 3Ј-OH primer concentration ( Fig. 2C) with the observed decrease of K m values in dependence of RP-A concentration (Fig. 2D) showed that [RP-A] 50 , the RP-A concentration giving half of the maximal decrease, was 41 nM, a value very close to the K m of pol ␣ for the 3Ј-OH primer, which is 39 nM.
RP-A Increases the Ability of pol ␣ to Discriminate between Correct and Incorrect Nucleotide Incorporation-Next, the effect of RP-A on the ability of pol ␣ to incorporate a wrong nucleotide was tested. It is known that the misincorporation efficiency of a pol is influenced by the nature of the mismatch resulting from the misalignment of the template-encoded base and the incoming nucleotide (43). Thus, to directly compare the effects of RP-A on the ability of pol ␣ to incorporate a wrong nucleotide, the homopolymeric substrate poly(dA)/oligo(dT) was used, which contains only adenines as template. As shown in Fig. 3A, the incorporation of radioactively labeled dTTP catalyzed by pol ␣ on such a template can be inhibited by the addition of unlabeled dNTPs as competitors. Each individual dNTP inhibited the reaction with different potencies, as indicated by the different K i values ( Table I) Table I. pol ␣ discriminates 330-fold against the C-A mismatch, but only 60-fold against the A-A and 30-fold against the G-A misincorporations, respectively. Thus, pol ␣ showed different misincorporation efficiencies for the resulting mismatches, in the order G-A Ͼ (i.e. more efficiently generated) A-A Ͼ C-A. When similar experiments were performed in the presence of RP-A, the misincorporation efficiencies for the G-A and A-A mismatches were reduced to about the level observed for the C-A misincorporation (see below and Table I). Fig. 3B shows the effect of different amounts of RP-A on the inhibition by dGTP of dTTP incorporation catalyzed by pol ␣ on poly(dA)/oligo(dT). The plot of the increase in the K i value of dGTP versus RP-A concentra-

Substrate
Competitor All of the values for the kinetic constants were calculated by computer simulation as described under "Materials and Methods." Standard deviations (ϮS.D.) were in all cases ՅϮ10%. Values reported are the mean of three independent experiments. The significance of the differences between the mean values obtained from reactions with and without RP-A were tested by a Student's t test under the null hypothesis that the true mean values were equal in all cases. The probability value that the reference hypothesis was true was p ՅϮ 0.05, and thus the observed differences were considered statistically significant. c n.d., not determined. tion showed a typical saturation kinetics (Fig. 3C), with a half-maximal stimulatory RP-A concentration (or [RP-A] 50 ) of 47 nM, very close to the value observed for the 3Ј-OH primer binding stimulation (Fig. 2D). Maximal increase of the K i values for dGTP inhibition was observed at RP-A concentrations close to the concentration of 3Ј-OH. Similar results were obtained when dATP inhibition was tested (data not shown). Sequencing Gel Analysis of the Products of Nucleotide Incorporation Catalyzed by pol ␣ on the d24:d66-mer Template-Because the template sequence was known, it was possible to force pol ␣ to make specific mismatches by adding the appropriate combinations of nucleotides in a reaction mixture. Fig. 4 shows the products of a reaction containing the heteropolymeric deoxyoligonucleotide resolved by denaturing polyacrylamide gel electrophoresis. The addition of different combinations of nucleotides generated strong pausing sites at the positions immediately preceding the mismatch, suggesting that incorporation of a wrong nucleotide was rate-limiting with respect to the elongation of a mismatched primer. For example, in the presence of the first encoded nucleotide, dCTP, a strong signal was detected at position ϩ1 (lane 2), whereas the addition of the first two nucleotides, dCTP and dTTP, resulted in the generation of a strong band at position ϩ2 (lane 3). Significant misincorporation products were detected with all of the combinations used (lanes 2-4), confirming the relatively low fidelity of pol ␣ in DNA synthesis.

RP-A Decreases the Amount of Misincorporated Products
Synthesized by pol ␣-The same reactions were performed in the presence of RP-A (Fig. 4, lanes 5-8). The first observation was that RP-A stimulated the correct incorporation of nucleotides by pol ␣ as judged by the increasing intensities of the bands corresponding to complementary nucleotide incorpora-tion (Fig. 4, compare positions 0, ϩ1, and ϩ2 in lanes 2-4 with the same positions in lanes [5][6][7][8]. In particular, the band at position 0 decreased significantly; this was expected, because RP-A imposed a block to misincorporation, which forced pol ␣ to elongate mainly the d24-mer primer to a d25-mer product but did not allow further elongation because of the lack of the complementary correct nucleotides required. The intensities of the bands were quantified by densitometric analysis, and the relative amounts of synthesized products at each position along the template were calculated as described in the figure legend, both in the absence and in the presence of RP-A. As shown in Fig. 5, with the different combinations of nucleotides tested, RP-A significantly decreased the amount of products generated by elongation of a mismatched primer by pol ␣, thereby increasing the accumulation of products at the position immediately preceding the misincorporation site (compare panel A with B in Fig. 5).

RP-A Influences Both the K m and the V max of Incorrect but Not of Correct Nucleotide Incorporation Catalyzed by pol ␣-To
investigate more closely the mechanism by which RP-A decreases the misincorporation efficiency of pol ␣, a detailed kinetic analysis of correct versus incorrect single nucleotide incorporation was performed. Reactions were carried out as shown in Fig. 4, lanes 2 and 6, respectively, but in the presence of different concentrations of dCTP. Quantification of the products at position ϩ1 (C-G base pair) and ϩ2 (C-A mismatch) at each nucleotide concentration allowed the calculation of the K m and V max values for the correct and incorrect incorporation reactions, as well as the specificity constant V max /K m . The computed values are listed in Table II. RP-A specifically decreased both the affinity and the reaction velocity for the misincorporation reactions, whereas it only slightly stimulated the correct nucleotide incorporation (ϳ2-fold increase in the V max /K m value). A comparison of the ratio between the V max and the K m values for correct and incorrect nucleotide incorporation in the absence and the presence of RP-A showed that pol ␣ alone discriminated ϳ540-fold against the C-A mismatch, a value comparable with the one derived with poly(dA)/oligo(dT) ( Table I). This value was increased to more than 3300-fold by the addition of RP-A, thus reducing the misincorporation more than 6-fold (Table II).

Characterization of the Functional Interaction between RP-A
and pol ␣-Primase-Several lines of evidence suggested that RP-A makes close contacts with both pol ␣ and the 3Ј-OH primer end. UV-cross-linking experiments showed the p70 subunit of RP-A bound to the single-stranded DNA of a primertemplate junction, whereas the p32 subunit was cross-linked very close to the 3Ј-OH primer end (33)(34)(35)(36). It has been shown by mutagenesis experiments that pol ␣ binds to the p70 and p32 subunits of RP-A. In particular, it was shown that the N-terminal half of the DNA-binding domain of the p70 subunit (amino acids 170 -327) was responsible for the enhancement of pol ␣ processivity and that the ssDNA binding activity of RP-A was required to achieve this stimulation (19). To investigate the molecular mechanisms and the functional significance of the RP-A/pol ␣ interaction, we examined in detail the effect of RP-A on pol ␣ catalytic activity. Our results clearly showed that RP-A was able to stimulate pol ␣ activity on different DNA templates (Fig. 1A). Under the assay conditions used, maximal stimulation was always achieved with an RP-A concentration equal to the 3Ј-OH primer ends concentration (Fig. 1B). When the effect of RP-A on the DNA binding affinity of pol ␣ was studied, the results showed that RP-A could increase the primer binding activity of pol ␣ (Fig. 2, A and B). Interestingly, the amount of RP-A required for half-maximal stimulation was almost identical to the concentration of 3Ј-OH primer ends required to half-saturate the pol ␣ active site (Fig. 2, C and D). These results might suggest that binding of RP-A leads to the formation of a ternary pol ␣⅐3Ј-OH primer⅐RP-A complex, with 1:1:1 stoichiometry and with increased stability with respect to the binary pol ␣⅐3Ј-OH primer complex. Stimulation of primer binding affinity was observed only in the presence of stoichiometric amounts of RP-A to the 3Ј-OH primer. An excess of RP-A suppressed this effect. This observation is consistent with the results of Braun et al. (19), who showed that processivity stimulation of pol ␣ by RP-A was restricted to subsaturating amounts of RP-A. Thus, the stoichiometry of RP-A binding to the pol ␣⅐3Ј-OH primer complex is crucial for its function. UV-cross-linking studies with RP-A and photoreactive primer templates showed that the labeling pattern of p70 and p32, and consequently their interaction with the template-primer junction, was strongly dependent on the ratio of RP-A to template concentration (36). A conformational change, involving repositioning of the p32 and p70 subunits, has been proposed to occur when saturating amounts of RP-A molecules associate during cooperative DNA binding. It is possible that, under such conformation, RP-A can no longer stabilize the pol ␣⅐3Ј-OH primer complex. Thus, transiently limiting RP-A binding to stoichiometric amounts with respect to the available template-primer junctions in the presence of pol ␣ might represent a novel mechanism for regulation of the early events in DNA synthesis. RP-A was not found to influence the polymerization rate of pol ␣, but the observed increase in primer binding stability could nevertheless account for the increase in pol ␣ processivity described by Braun et al. (19). A major difference among pol ␣ and the other two replicative pols, pol ␦ and pol ⑀, is the lack of an intrinsic (or functionally associated) 3Ј3 5Ј proofreading exonuclease (3). As a result, the overall misincorporation rate of pol ␣ has been estimated at about 2.5 ϫ 10 Ϫ4 (44), approximately 5-20-fold higher than for pol ␦ and pol ⑀. The RP-A-dependent change in the molecular structure of the pol ␣⅐3Ј-OH primer complex could potentially influence the ability of pol ␣ to tolerate a misaligned nucleotide within its active site (45,46). Indeed, the fidelity of DNA synthesis catalyzed by pol ␣ when part of a multiprotein replication complex was found to be higher than that of purified pol ␣, suggesting a role for accessory factors (47,48). Using a combination of in vitro DNA synthesis and genetic screening for detecting mutations in Escherichia coli, RP-A has been found to increase by 5-6-fold the fidelity of pol ␣, even if no explanation for this effect at the molecular level was presented (49). However, another study with a different approach failed to detect a specific role of RP-A in modulating pol ␣ fidelity (50). In the present study, when the effect of RP-A on the ability of pol ␣ to incorporate wrong nucleotides was analyzed in in vitro assays with purified proteins, the results showed that the overall misincorporation efficiency was reduced by about 5-6-fold (Figs. 3 and 5 and  Tables I and II). A kinetic analysis showed that this effect was driven by RP-A physical association with pol ␣ and the 3Ј-OH Error bars indicate the errors (ϮS.D.) calculated over three independent experiments. The significance of the differences between the mean values obtained from reactions with and without RP-A were tested by a Student's t test under the null hypothesis that the true mean values were equal in all cases. The probability value that the reference hypothesis was true was p Յ 0.05, and thus the observed differences were considered statistically significant. primer end, which reduced both the affinity and the catalytic efficiency of pol ␣ for incorrect versus correct nucleotide incorporation (Table II). Thus, the ternary pol ␣⅐3Ј OH primer⅐RP-A complex, besides having an increased stability, also showed a reduced ability to tolerate misalignments between the DNA template and the incoming nucleotide.
RP-A as a Fidelity Clamp-The role of the auxiliary proteins in DNA replication is to provide pols with particular properties, which make them better suited to perform the difficult task of replicating the whole cellular genome (9). For example, highly processive DNA synthesis is achieved by replicative pols through their interaction with a class of auxiliary proteins called sliding clamps or processivity clamps. In eukaryotic cells, such a role is fulfilled by the accessory protein PCNA, which acts as a processivity factor for both pol ␦ and pol ⑀, increasing their primer binding efficiency. The presence of a strongly evolutionary conserved processivity factor for pol ␦ is consistent with its role in the DNA replication process, namely elongation of leading and lagging strands. On the other hand, both pol ␦ and pol ⑀ possess a 3Ј3 5Ј proofreading exonuclease activity, and thus they should not require any fidelity-enhancing factor. pol ␣ performs a different task in DNA replication, synthesizing the short (30 nucleotides) RNA/DNA primers required for subsequent processive DNA synthesis by pol ␦ (or pol ⑀) (4). The switch between primase and pol activity, leading to the RNA-to-DNA synthesis transition, occurs through an intramolecular mechanism that does not require dissociation of pol ␣ from the primer (7). Additionally, the lack of any intrinsic proofreading function for pol ␣ could lead to misincorporation events during primer synthesis (44). The results presented here indicate that the RP-A-template-primer complex is a better substrate for pol ␣, allowing more processive and accurate synthesis of the RNA/DNA primers. Under this respect, RP-A behaves differently from the sliding clamps PCNA and gp45, which have been shown to reduce the accuracy of the cognate pols, allowing these enzymes to more easily replicate across DNA lesions and to extend mispaired primer ends (51,52). Thus, RP-A might represent the first example of a novel class of auxiliary factors, which we propose to call "fidelity clamps." Recent data suggested that a preinitiation complex between RP-A, Cdc45p, Mcm4p, and pol ␣ is formed at the replication origins (41,42). Moreover, pol ␣/RP-A interaction is also important in the switch between pol ␣ and pol ␦ at the lagging strand, which is dependent on replication factor C and PCNA (5,53). Thus, it appears that a continuous interaction between RP-A and pol ␣-primase is maintained throughout the early steps of DNA replication. a All of the values for the kinetic constants were calculated by computer simulation as described under "Materials and Methods." Standard deviations (ϮS.D.) are indicated. Values reported are the mean of three independent experiments. Significance of the differences between the mean values obtained from reactions with and without RP-A were tested by a Student's t test under the null hypothesis that the true mean values were equal in all cases. The probability value that the reference hypothesis was true was p Յ 0.05, and thus the observed differences were considered statistically significant.